We have found that lymphocytes and IFNgamma function as an effective cancer immunosurveillance system to protect mice against development of spontaneous and chemically induced primary tumors. However, we also found that tumors arising in immunodeficient mice are more immunogenic than those from immunocompetent mice indicating that the immune system also selects for tumor variants that express reduced immunogenicity. These observations led us to refine the cancer immunosurveillance hypothesis into the "Cancer Immunoediting Hypothesis" that stresses the paradoxical host-protective and tumor sculpting effects of immunity on developing tumors. Whereas the Cancer Immunoediting hypothesis was formulated on the basis of functional criteria (different in vivo growth phenotypes of tumors from immunodeficient versus wild type mice), we now seek to define the process at the molecular level. Using gene-profiling approaches, we recently identified 180 genes that are differentially expressed in highly immunogenic, unedited sarcomas from immunodeficient mice versus edited sarcomas from wild type mice. CD1d represents one of these genes and its expression is selectively down regulated in edited sarcomas that form in the presence of an intact immune system. Injection of an unedited, highly immunogenic tumor (F535) that expresses high CD1d levels into partially immunodeficient mice leads to formation of escape variants that (a) display significantly reduced levels of CD1d mRNA and protein and (b) grow progressively when injected into unmanipulated wild type mice. Enforced expression of CD1d in F535 escape variants restores their high immunogenicity. Capitalizing on these novel observations, we will now define the molecular targets and underlying mechanisms of cancer immunoediting by pursuing the following four specific aims. In Specific Aim 1 we will determine whether CD1d is a generalized target of the cancer immunoediting process. In Specific Aim 2 we will identify and validate other genes that are additional targets of cancer immunoediting giving particular emphasis to our recent discovery that this process may additionally target the pathway leading to production of the glycolipid ligands of CD1d. In Specific Aim 3 we will define the mechanism(s) underlying the silencing of these genes. In Specific Aim 4 we will identify the pathway "editors" giving special emphasis to assessing a role for NKT cells in the editing process. This work will provide the first molecular and mechanistic insights into the cancer immunoediting process and may also establish molecular guidelines to judge the extent to which a tumor has been edited.